DESCRIPTION: A higher performance transmit-receive (T/R) switch is needed for high field, 3T and 4T NMR systems now being delivered and applied to human biomedical research. The increases in signal-to-noise, spectral resolution, and other proven advantages of high field NMR point toward the need to develop better optimized high field (3T-6T) systems. Other than the high field magnet itself, a key technological difference between the current 1.5T clinical systems and higher field systems for human studies is found in the RF front end of the system. When compared to lower field clinical systems, higher RF peak power transmit levels are required for human studies at higher fields to cover broader spectral bandwidths, correct for chemical shift dispersion error, and to a lesser extent compensate for RF shielding and losses in human tissues. These higher RF power levels and higher frequencies challenge the performance of the conventional quarter wave PIN switch designs often used for lower frequency and lower peak power applications. Many of the present switch designs are limited in both peak and average power handling capability, and in bandwidth. The balanced duplexer design is proposed to solve these problems. The balanced duplexer design used in radar is inherently a higher power, higher speed, and less noisy switch for high power and high frequency applications. Additionally, and unlike the tuned quarter wave switch designs most often used in the MRI industry, the switch proposed is broad banded for high speed, flexible and convenient use in multi-nuclear studies common at high fields. We propose to test the feasibility of using this T/R switch for high field human applications by building a prototype switch which meets or exceeds a list of desired performance criteria listed as specific aims. The Phase II plan is to make this switch available in product form to independent investigators, and to the industry.